paper-with-me

홈 › Papers

LatentPrompt: Optimizing Promts in Latent Space

2025-08-04 · Mateusz Bystroński, Grzegorz Piotrowski, Nitesh V. Chawla, Tomasz Kajdanowicz arxiv

Recent advances have shown that optimizing prompts for Large Language Models (LLMs) can significantly improve task performance, yet many optimization techniques rely on heuristics or manual exploration. We present LatentPrompt, a model-agnostic framework for prompt optimization that leverages latent semantic space to automatically generate, evaluate, and refine candidate prompts without requiring hand-crafted rules. Beginning with a set of seed prompts, our method embeds them in a continuous latent space and systematically explores this space to identify prompts that maximize task-specific performance. In a proof-of-concept study on the Financial PhraseBank sentiment classification benchmark, LatentPrompt increased classification accuracy by approximately 3 percent after a single optimization cycle. The framework is broadly applicable, requiring only black-box access to an LLM and an automatic evaluation metric, making it suitable for diverse domains and tasks.

📄 PDF Abstract BibTeX arXiv:2508.02452

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Combining Latent Space and Structured Kernels for Bayesian Optimization over Combinatorial Spaces

2021-11-01 · NeurIPS 2021 12 · Aryan Deshwal, Janardhan Rao Doppa

We consider the problem of optimizing combinatorial spaces (e.g., sequences, trees, and graphs) using expensive black-box function evaluations. For example, optimizing molecules for drug design using physical lab experim…

Bayesian OptimizationDrug DesignInductive Bias

Generative Co-Design of Antibody Sequences and Structures via Black-Box Guidance in a Shared Latent Space

2025-08-15 · Yinghua Yao, Yuangang Pan, Xixian Chen arxiv

Advancements in deep generative models have enabled the joint modeling of antibody sequence and structure, given the antigen-antibody complex as context. However, existing approaches for optimizing complementarity-determ…

From Parameters to Behaviors: Unsupervised Compression of the Policy Space

2025-09-26 · Davide Tenedini, Riccardo Zamboni, Mirco Mutti, Marcello Restelli arxiv

Despite its recent successes, Deep Reinforcement Learning (DRL) is notoriously sample-inefficient. We argue that this inefficiency stems from the standard practice of optimizing policies directly in the high-dimensional …

Reinforcement LearningContinuous Control

Optimizing Latent Space Directions For GAN-based Local Image Editing

2021-11-24 · Ehsan Pajouheshgar, Tong Zhang, Sabine Süsstrunk

Generative Adversarial Network (GAN) based localized image editing can suffer from ambiguity between semantic attributes. We thus present a novel objective function to evaluate the locality of an image edit. By introduci…

DisentanglementGenerative Adversarial Network

Efficient Test-Time Adaptation through Latent Subspace Coefficients Search

2025-10-13 · Xinyu Luo, Jie Liu, Kecheng Chen, Junyi Yang 외 arxiv

Real-world deployment often exposes models to distribution shifts, making test-time adaptation (TTA) critical for robustness. Yet most TTA methods are unfriendly to edge deployment, as they rely on backpropagation, activ…

Test-time Adaptation